Instructions to use QuantFactory/Meta-Llama-3-225B-Instruct-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use QuantFactory/Meta-Llama-3-225B-Instruct-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="QuantFactory/Meta-Llama-3-225B-Instruct-GGUF",
	filename="Meta-Llama-3-225B-Instruct.Q2_K-00001-of-00003.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use QuantFactory/Meta-Llama-3-225B-Instruct-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf QuantFactory/Meta-Llama-3-225B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf QuantFactory/Meta-Llama-3-225B-Instruct-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf QuantFactory/Meta-Llama-3-225B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf QuantFactory/Meta-Llama-3-225B-Instruct-GGUF:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf QuantFactory/Meta-Llama-3-225B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf QuantFactory/Meta-Llama-3-225B-Instruct-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf QuantFactory/Meta-Llama-3-225B-Instruct-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf QuantFactory/Meta-Llama-3-225B-Instruct-GGUF:Q4_K_M

Use Docker

docker model run hf.co/QuantFactory/Meta-Llama-3-225B-Instruct-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use QuantFactory/Meta-Llama-3-225B-Instruct-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "QuantFactory/Meta-Llama-3-225B-Instruct-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "QuantFactory/Meta-Llama-3-225B-Instruct-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/QuantFactory/Meta-Llama-3-225B-Instruct-GGUF:Q4_K_M

Ollama
How to use QuantFactory/Meta-Llama-3-225B-Instruct-GGUF with Ollama:
```
ollama run hf.co/QuantFactory/Meta-Llama-3-225B-Instruct-GGUF:Q4_K_M
```

Unsloth Studio new

How to use QuantFactory/Meta-Llama-3-225B-Instruct-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for QuantFactory/Meta-Llama-3-225B-Instruct-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for QuantFactory/Meta-Llama-3-225B-Instruct-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for QuantFactory/Meta-Llama-3-225B-Instruct-GGUF to start chatting

Docker Model Runner
How to use QuantFactory/Meta-Llama-3-225B-Instruct-GGUF with Docker Model Runner:
```
docker model run hf.co/QuantFactory/Meta-Llama-3-225B-Instruct-GGUF:Q4_K_M
```

Lemonade

How to use QuantFactory/Meta-Llama-3-225B-Instruct-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull QuantFactory/Meta-Llama-3-225B-Instruct-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Meta-Llama-3-225B-Instruct-GGUF-Q4_K_M

List all available models

lemonade list

Meta-Llama-3-225B-Instruct

This is quantized version of mlabonne/Meta-Llama-3-225B-Instruct created using llama.cpp

Meta-Llama-3-225B-Instruct is a self-merge with meta-llama/Meta-Llama-3-70B-Instruct.

It was inspired by large merges like:

I don't recommend using it as it seems to break quite easily (but feel free to prove me wrong).

🧩 Configuration

slices:
- sources:
  - layer_range: [0, 20]
    model: mlabonne/Meta-Llama-3-120B-Instruct
- sources:
  - layer_range: [10, 30]
    model: mlabonne/Meta-Llama-3-120B-Instruct
- sources:
  - layer_range: [20, 40]
    model: mlabonne/Meta-Llama-3-120B-Instruct
- sources:
  - layer_range: [30, 50]
    model: mlabonne/Meta-Llama-3-120B-Instruct
- sources:
  - layer_range: [40, 60]
    model: mlabonne/Meta-Llama-3-120B-Instruct
- sources:
  - layer_range: [50, 70]
    model: mlabonne/Meta-Llama-3-120B-Instruct
- sources:
  - layer_range: [60, 80]
    model: mlabonne/Meta-Llama-3-120B-Instruct
- sources:
  - layer_range: [70, 90]
    model: mlabonne/Meta-Llama-3-120B-Instruct
- sources:
  - layer_range: [80, 100]
    model: mlabonne/Meta-Llama-3-120B-Instruct
- sources:
  - layer_range: [90, 110]
    model: mlabonne/Meta-Llama-3-120B-Instruct
- sources:
  - layer_range: [100, 120]
    model: mlabonne/Meta-Llama-3-120B-Instruct
- sources:
  - layer_range: [110, 130]
    model: mlabonne/Meta-Llama-3-120B-Instruct
- sources:
  - layer_range: [120, 140]
    model: mlabonne/Meta-Llama-3-120B-Instruct
merge_method: passthrough
dtype: float16

💻 Usage

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "mlabonne/Meta-Llama-3-220B-Instruct"
messages = [{"role": "user", "content": "What is a large language model?"}]

tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    torch_dtype=torch.float16,
    device_map="auto",
)

outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])